Re: git: fcf990eaf36a - main - audio/sayonara: enable PulseAudio, repair Python scripts

On Wed, Jun 12, 2024 at 11:56:17AM +0300, Gleb Popov wrote:
> ...
> Like I said before, take any upstream project you like, implement the
> native OSS support for it, persuade upstream to accept your patches
> and then enjoy your PulseAudio-free system.

I've been enjoying full-featured PulseAudio-free FreeBSD desktop since,
like, forever.  Which patches are lacking that prevent you from doing
the same?

> But we both know it is unrealistic

*sigh* I don't know how to make it more clear, but one does not have
to run Lennart's crapware on her desktop to fully enjoy all multimedia
features for music, gaming, YouTube, or streaming.  We traditionally
kept PA support in ports OFF by default for that very reason.  I could
imagine some rare use case where it might be required, or when someone
got used to it coming to FreeBSD from GNU/Linux, but it had always been
opt-in and should stay this way.

> then why keep ranting on the mailing list?

I can't agree that a reply to some commit message is ranting.

> Unfortunately, libinotify-kqueue isn't that thin as you'd think [...]

Perhaps so, but before we convince or help developers to start using
our native APIs, we still need those shims.  This is a sharp contrast
to PA which you can delete from your system or even the ports tree and
nothing would really break.

./danfe